One of the biggest discussions in the visual effects world is the increasing use of artificial intelligence technologies by production studios, but the technology is far older, far broader and far more widely used than many might expect.
Whilst it is most commonly associated with more recent technological advances such as the use of generative AI and large language models to produce some types of visual effects, AI has a lot of broad uses within the industry.
Were it not for an early application of AI visual effects technology, one of the most ambitious and successful film series would not have been possible, and the effect it has had on film as a whole is dramatic.
One Ring To Rule Them All
Whilst there have been many books that have been described as “unfilmable” in the history of cinema, arguably the most famous example with the most attempts has been J.R.R. Tolkien’s The Lord of The Rings.
Because of the scale of the books and particularly the scale of many of the most pivotal battles in the novel, as well as the perception in the United States that the books had only niche popularity, the only major adaptation of the books was the 1978 animated adaptation by Ralph Bakshi.
He also struggled to capture the scale of the books and the major set pieces into one film, relying heavily on the rotoscoping animation technique, a form of pre-computer graphics motion capture where the motions of real actors were traced, as well as a range of visual filters and unusual effects.
It proved there was an appetite for a more complete Lord of the Rings adaptation, but it would take until another relative outsider in Peter Jackson explored what was possible with a mix of evocative landscapes, advances in model design, judicious use of visual effects, and assistance from AI.
A MASSIVE Change
The Lord of The Rings Trilogy would end up finding the answers to many visual effects questions, most notably involving the entirely motion-captured performance of Andy Serkis’ Gollum/Smeagol, but the most fundamental issue involved the massive set piece battles.
The Battle of Helm’s Deep in The Two Towers, and both the Battle of the Pelennor Fields and Battle of the Morannon in The Return Of The King involved thousands of combatants, something that would be impossible to execute using extras and would be time-consuming to animate manually.
The solution was the MASSIVE crowd simulation software tool, developed in the late 1990s and adapted for use by the Lord of the Rings production team.
It worked by developing a series of relevant computer-generated agents, who could move and perform trained actions realistically, typically by being loaded with motion capture or animation files.
They would generate soldiers based on existing sets of armour and clothing, as well as sets of animations, and then the software would choose via fuzzy logic realistically different animations that created the impression of vast swathes of combatants.
It worked so effectively and seamlessly because, at Peter Jackson’s insistence, most of the visual effects were practical, relying on model shots, prosthetics and beautiful location shoots alongside CGI characters, composite effects, extensive use of colour grading and digital matte paintings.
